Probability of Failure for Slopes with Sensitivity Analysis

Abstract

The significance of random variables on the probability of failure of slopes is generally evaluated by changing a different set of values for each variable and repeating the approach several times to conclude a trend. This method is very straightforward, but requires time and computational resources; and it is inefficient especially when the number of random variables is large. This paper presents an analysis approach in which the sensitivity analysis is introduced to quantify the influence of the random variables on the probability of failure for slopes. In this study, the factor of safety of the slope was determined using the ordinary method of slices, in which the Mohr Coulomb strength parameters were considered to be independent random variables. The probabilistic analysis was performed using the First-Order-Reliability method to search for the surface with the minimum reliability index, β, representing the most probable failure surface. Based on β computed from the previous probabilistic analysis, a sensitivity analysis was conducted to identify the significance of the soil strength parameters on the probability of failure for both the deterministic slip surface (min. FS) and the probabilistic slip surface (min. β).